Stereocontrolled total synthesis of (.+-.)- and (+ ... - ACS Publications

Natural Products Synthesis: Enabling Tools To Penetrate Nature's Secrets of Biogenesis and Biomechanism. Robert M. Williams. The Journal of Organic ...
7 downloads 0 Views 371KB Size
J . A m . Chem. SOC.1984, 106, 5748-5750

5148

that spectrophotometry cannot be used for identification of the site of reduction in these synthetic chlorins and in those from natural sources. Finally, it should be mentioned that this procedure offers access to a number of isomeric tetrapyrrole compounds suitable for spectroscopic study of model biological systems containing green hemes. Moreover, previous total syntheses of 2-vinylrhodoporphyrin XV22and 2,4-divinylrhodoporphyrinXV,I5 coupled with the published transformation of 2-vinylrhodochlorin into chlorophyll a,I4 open up a viable route for the efficient total synthesis of both chlorophyll a and 2,4-divinylchlorophylla, the latter having been the topic of considerable attention in recent times.23 This work is currently in progress.

Acknowledgment. This research was supported by grants from the National Science Foundation (CHE-81-20891) and the National Institutes of Health (HL 22252). We thank Frank Bobe for carrying out the N O E study. Supplementary Material Available: Proton N M R spectra (360 MHz), melting points, and electronic absorption spectra of compounds 10 and 12-15 and the electronic absorption spectrum of the iron(II1) chloride complex of 12 (typical example) (2 pages). Ordering information is given on any current masthead page. (22) Howarth, T. T.; Jackson, A. H.; Kenner, G. W. J . Chem. SOC.,Perkin Trans. 1 1974,502-5 11. (23) Rebeiz, C. A. Chem Tech, 1982, 12, 52-63. Castelfranco, P. A.; Beale, S. Annu. Rev. Plant Physiol. 1983,34, 241-278.

Stereocontrolled Total Synthesis of (&)- and (+)-Bicyclomycin: New Carbon-Carbon Bond-Forming Reactions on Electrophilic Glycine Anhydride Derivativest Robert M. Williams,** Robert W. Armstrong, and Jen-Sen Dung

Department of Chemistry, Colorado State University Fort Collins, Colorado 80523 Received May 29, 1984 Bicyclomycin’ (1) is a novel antibiotic that is biosynthetically derived2 by the oxidative cyclodimerization of the amino acids leucine and isoleucine. Bicyclomycin has recently achieved

,‘\OH

2 -

3 -

commerical stature3 on a worldwide basis as a clinically useful antibiotic and is now produced on large scale from cultures of Streptomyces sapporonensis. ~~~

‘Dedicated to the memory of the late Professor Kunio Sakan. * N I H Research Career Development Awardee 1984-1989. ( I ) Miyoshi, T.; Miyairi, N.; Aoki, H.; Kohsaka, M.; Sakai, H.; Imanaka, H . J . Antibior. 1972,25, 569. See ref 4 for additional citations for isolation, structure determination, and isolation from Streptomyces aizunensis. (2) (a) Miyoshi, T.; Iseki, M.; Konomi, T.; Imanaka, H.J . Antibiot. 1980, 33, 480. (b) Miyoshi, T.; Iseki, M.; Konomi, T.; Imanaka, H. Zbid. 1980,33, 488. (3) ’Merck Index”, 10th Ed.; Merck: Rahway, NJ, 1983; no. 1213. Bicozamycin is the commercial synonym (Fujisawa Pharmaceutical Company, Ltd., Japan) for bicyclomycin (aizumycin).

0002-7863/84/ 1506-5748$01 .50/0

We have recently reported4 the synthesis and regiocontrolled bridgehead carbanion elaboration of the 4-demethylene nucleus 2. In order to reduce this efficient model study5 to a total synthesis6 of 1, two difficult problems had to be addressed: (1) introduction of the C4-C5 exo-methylene moiety via a suitably oxidized isoleucine precursor and (2) selection of a suitable blocking group for the amides. In this paper, we wish to report a completely regio- and stereocontrolled total synthesis of bicyclomycin from the nucleus 3 that features a fundamentally new and generally useful C - C bond-forming reaction via electrophilic coupling to a glycine anhydride derivative. As shown in Scheme I, 1,4-bis(p-methoxybenzyl)- and 1,4dibenzyl-2,5-piperazinedionewere brominated’ and condensed with the sodium salt of 2-mercaptopyridine (THF, 25 OC, 30 min) to afford the crystalline syn-bis(su1fide) 5. Precomplexation of 5 with 1 equiv of silver (I) triflate in T H F at 25 OC for 10 min followed by addition of 1 equiv of butyrolactone trimethylsilyl enol ether (2 h, 25 “C) furnished the lactones 6 (1.3:1, syn:anti; 1.8:l ratio, epimeric at the lactone a-carbon) in 7 1 % yield.6 It turned out to be critical to precomplex 5 with the silver salt before addition of the nucleophile to effect coupling. We were quite surprised to find that the silver complex of 5 is indefinitely stable in solution (THF, CH2Cl2,CHC13) and cleanly reacts, producing 6 upon addition of the trimethylsilyl ketene acetal. Additionally, the reaction proceeds predominantly with overall retention of stereochemistry with respect to the departing thiopyridyl residue and the newly attached lactone moiety. An X-ray crystallographic analysis9 of the major syn diastereomer 6a established the relative configuration (shown). Most importantly, we found that the product 6 completely resists further C-C substitution at the remaining thiopyridyl residue (excess AgOTf/ketene silyl acetal) at C-3 so that absolutely no 3,6-biscoupled products are observed. This remarkable chemoselectivity is highly significant since a major competing side reaction observed in the nucleophilic C-functionalization of N-substituted glycine anhydride enolates (Le., of 4) is 3,6-dis~bstitution.~~ Reduction of the major syn and anti lactones 6 afforded the diol 7, which was cleanly cyclized1° to the desired bicyclic alcohol 8 in the presence of silver(1) triflate in T H F at 25 OC. Dehydration of 8 to the bicyclic olefin 9 was readily accomplished in three steps (Scheme I, steps e, f, g).

(4) (a) Williams, R. M.; Anderson, 0. P.; Armstrong, R. W.; Joey, J.; Meyers, H.; Eriksson, C. J . Am. Chem. SOC.1982,104, 6092. (b) Williams, R. M.; Dung,J.-S.; Josey, J.; Armstrong, R. W.; Meyers, H.Zbid. 1983,105, 3214. (c) Armstrong, R. W.; Dung, J.-S.; Williams, R. M. “Abstracts of Papers”, 185th National Meeting of the American Chemical Society, Seattle, WA, March 1983; American Chemical Society: Washington, DC, 1983; O R G N 10. (5) For other synthetic approaches to 1, see the references cited in ref 4; see also: (a) Yates, P.; Hoare, J. H . Can. J . Chem. 1983,61, 519. (b) Yates, P.; Hoare, J. H. Ibid. 1983, 61, 1397. (c) Dirlam, J. P.; James, R. B.; Shoop, E. V . ‘Abstracts of Papers“, 185th National Meeting of the American Chemical Society, Seattle, WA, March 1983; American Chemical Society: Washington, DC, 1983; ORGN 9. (6) For a recent total synthesis, see: (a) Nakatsuka, S.; Yamada, K.; Yoshida, K.; Asano, 0.;Murakami, Y.; Goto, T. Tetrahedron Left. 1983,24, 5627. (b) Nakatsuka, S.; Goto, T . Heterocycles 1984,21, 61. (7) Trown, P. W. Biochem. Biophys. Res. Commun. 1968,33, 402. (8) The coupling reaction of 5a to afford 6a proceeded to give a 2:l ratio of syn lactones. The major syn diastereomer was directly converted to Sa by LiAIH, reduction and cyclization. The minor syn lactone could either be epimerized to the major syn diastereomer or converted to 913as described in ref IO. (9) Williams, R. M.; Armstrong, R. W.; Maruyama, L. K.; Dung, J.-S.; Anderson, 0. P., unpublished results. (10) The minor syn lactone 6b could be epimerized to a 1:l mixture of the two syn diastereomers (0.1 N NaOH, THF, 25 “C) or reduced to the corresponding diol (LiAlH,). This diol was converted to the desired bicyclic system through ( I ) selective silylation at the 3”-hydroxyl (Me,Bu+SiCI, DMAP, Et3N, CH,CI,), (2) mesylation (MsC1, Et,N, THF), and (3) cyclization with C U ( C I O ~ ) ~ / T H25 F , “C, to afford the bicyclo[4.2.2] mesylate (epimeric at C-5, cf. structure 8) which was directly converted to olefin 9 (Scheme I, steps f and g).

0 1984 American Chemical Society

J . A m . Chem. SOC.,Vol. 106, No. 19, 1984 5149

Communications to the Editor Scheme I o

k

BnN$Bn 0

70 %

4 -

7I %

7 -

6 0%

92 %

82 %

8 -

9 -

h.i

*OJ

6 6%

IO

52 %

73%

95%

3 series, Bn=CH2Ph series , Bn = C H2ph-p-0 Me

xor O*Me

II , R = H

2 ,RzCOCF3

:agents and Conditions: (a) 2 equiv ofN-bromosuccinimide, catalytic benzoyl peroxide, CCI,, re I X , 30 m i n ; ( b ) 2.0 equiv of NaS(py), T H F , 25 "C, 30 min; (c) 1 equiv of AgOTf, T H F , 25 'C; (d) 1 equiv of LiAlH,, THF, 25 "C, 1 min then, Na,SO;lOH,O; (e) 2.5 equiv of methanesulfonyl chloride, Et,N ( 2 5 equiv), THF, 25 "C, 1 2 11; (f) 2.2 equiv of NaBH,SePh, THF, reflux, 2.2 h ; (8) 30% H,O, (5 equiv) THF, reflux, 20 min; (h) 1.5 equiv ofiz-BuLi, HMPA ( 2 equiv), (Me,N),P (2 equiv), THF, -78 "C, 1 min; (i) 0, (gas) 15 min, -100 "C; 0 ) 2.3 equiv of n-BuLi, T H F , -100 "C; (k) 10 equiv of (F,CCO)lO, 15 equiv of DMAP, CH,Cl,, 25 T, 20 min; (1) 4 equiv of ceric ammofiium nitrate, CH,CN/H,O (4:1,0.2 M), 25 "C, 35 min, t h e n silica gel PTLC (2070 MeOH in CHC1,). Q

Regioselective" bridgehead hydroxylation of 9 (Scheme I, steps h and i) afforded the desired bridgehead alcohol 10 (55%). Formation of the dianion4 of 10 followed by aldol condensation with (~)-2,2,4-trimethyl-1,3-dioxolane-4-carbo~aldehyde~~ afforded a single diastereomer 11 (95%); no evidence for the formation of any of the other three possible diastereoisomers on many trials could be obtained under these condition^.'^ We found it (1 1) We initally investigated the bridgehead functionalization of the terr-butyldimethylsilyl ether i obtained from 8a (C1SiMe2Bu', Et3N, CH2C12,

- ,Rl:H, IV

R2=SiMe2Bu t

-v ,R , = O H ,R2=SlMe2Bu

DMAP). Treatment of i with n-BuLi in THF/HMPA at -100 'C followed by quenching with CH31 afforded exclusively the methylated derivative ii. This regioselectivity is in contradistinction with that we have observed4 for 2. Regio- and stereoselective aldol condensation of the bridgehead carbanion 11 afforded a single of i (LDA/THF, -100 "C) as described for 10 diastereomer iii (80%). Silylation (Bu+Me,SiOTf, 2,6-lutidine, CH2CI,, 25 "C) of the secondary alcohol followed by hydroxylation (t-BuLi/THF -100 "C, O2 quenching) afforded the alcohol v (78%). (12) (a) Maag, H.; Blount, J. F.; Coffen, D. L.; Steppe, T. V.; Wong, F. J . Am. Chem. SOC.1978, 100, 6786. (b) See also: Dung, J.-S.; Armstrong, R. W.; Anderson, 0. P.; Williams, R. M. J . Org. Chem. 1983,. 48, 3592.

-

necessary to block the secondary hydroxyl group of llb to effect the clean removal14of the p-methoxybenzyl residues and preclude a competing rearrangement similar to that recently reported by Wacker.I5 Thus, treatment of llb with TFAA/DMAP in CH2C12 afforded the corresponding 1'-0-trifluoroacetate 12. Reaction of this material with 4 equiv of ceric ammonium nitrateI6 in acetonitrile/H20 (0.2 M) followed by PTLC on silica gel directly furnished totally synthetic bicyclomycin (3 1% overall from llb). The synthetic material was identical with the natural sample" by 'H NMR, 13CNMR, IR, MS, and TLC behavior. By carrying out the aldol condensation of racemic 10 with optically active (-)-2,2,4-trimethyl- 1,3-dioxolane-4-carboxaldehyde~2b (83% ee) and following exactly the same procedure as above, optically active bicyclomycin, [a]250 +50.8' (MeOH, 78% ee), was obtained; this constitutes the first total synthesis of bicyclomycin in optically active form. Thus, the total synthesis of bicyclomycin has been achieved in only 12 chemical steps with complete regio- and stereocontrol (4.6% yield overall). In addition, the new C-C (13) At -78 "C, a small amount of the (2-2' diastereomer can be isolated. The related aldol condensation in the Goto synthesis6 exhibited relatively poor diastereoselectivity (3:l:l :O ratio, 41% yield). (14) The N-benzyl series was terminated at this junction since all attempts to reductively remove the N-benzyl groups on any bicyclic derivative (i-vi, 813,2) failed to produce any quantity of the desired deprotected compounds; reductive cleavage of the C , - O ether linkage and saturation of the aromatic benzylic rings were the only types of reactivity observed. Conditions examined: 20% Pd/C, H2, 1 atm, EtOH, 80 OC; 20% PtO,, H2; 20% Pd(OH),/H,; a range of solvents, temperatures, and H2 pressures were examined for each catalyst; see also, ref 6. (15) Wacker, 0.; Kump, W.; Muller, B. W. Tetrahedron Lett. 1983, 24, 5607. . .. .

(16) The excellent procedure reported by Yoshimura was utilized: Yoshimura, J.; Yamaura, M.; Suzuki, T.; Hashimoto, H. Chem. Letr. 1983, 1001.

5750

J . Am. Chem. SOC.1984, 106, 5750-5751

coupling reaction we have developed for this synthesis represents a highly useful chemoselective method for preparing hitherto inaccessible a-C-homologated piperazinediones and, potentially, other a-amino acid derivatives9 The methodology described herein is uniquely adaptable to the preparation of many structurally diverse bicyclomycin analogues that cannot be prepared by modification of the abundantly available natural product nor from any of the other published516synthetic efforts. Finally, we have found that the p-methoxybenzyl groups can be cleanly and reliably removed from any of these bicyclic structures" (Le., 2 (R = CH2Ph-p-OCH3),8b, 9b, and lob) to afford the hydrophilic "free" amides. Biological and mechanistic studies utilizing this chemistry shall be reported in due course from these laboratories.

[M + H

Acknowledgment. We gratefully acknowledge the National Institutes of Health Grant ROlAIGM 18957 for financial support of this work. We thank Fujisawa Pharmaceutical Co., Ltd., Japan, for the generous gift of natural bicyclomycin used for comparison. N M R measurements at 360 MHz were obtained at the Colorado State University Regional N M R Center, funded by the National Science Foundation Grant C H E 78-18581.

__

-

2600

mlz

2800

Supplementary Material Available: Complete spectroscopic and analytical data for all new compounds (12 pages). Ordering information is given on any current masthead page. (1 7) Preliminary antimicrobial assays of totally synthetic (i)-bicyclomycin against E . coli 94 and Kkbsiella pneumoniae 369 show that the racemic material exhibits half the activity of the natural compound; numerous Ndeprotected bicyclic analogues have been evaluated for antimicrobial activity: Williams, R. M.; Armstrong, R. W.; Dung,J.-S., unpublished results. We thank Drs. Hans Maag and David Pruess of Hoffman La-Roche, Inc., for performing the assay.

Fast Atom Bombardment Mass Spectroscopy (FABMS) of Polyoxoanions

[M+ByN]' ~

____ 3400

sh ~~

3600

3800 m/z

Figure 1. (A) Positive-ion spectrum of H4SiWI2O4,. (B) Negative-ion spectrum of H4SiWl,0,. Note that molecular ion and the loss of 0 and WO, are observed in both spectra A and B. Not shown are the sequential

losses of WO, in both spectra and much smaller peaks above 3000 m / z due to the attachment of WO, fragments or of thioglycerol in the negative-ion spectrum. (C) Negative-ion spectrum of K4H,SiW9V3O4,. Extensive exchange of cations (cationization) is observed. Sequential loss of 0 and WO, (not shown) is also observed. (D) Positive-ion spectrum of (Bu4N),CpTi.SiW,V30,0. Only peaks corresponding to cation exchange and loss of 0 are observed. T h e sequential loss of WO, is not observed.

Richard G . Finke,*' Michael W. Droege,] J. Carter Cook,* and Kenneth S. Suslick**

Department of Chemistry, University of Oregon Eugene, Oregon 97403 School of Chemical Sciences University of Illinois at Urbana- Champaign Urbana, Illinois 61801 Received May 3, 1984 Polyoxoanion chemistry3 is a field poised for a rapid development with a wide range of potential application^.^^^^^^^ Ha mpering (1) University of Oregon. (2) University of Illinois. (3) (a) Pope, M. T. In "Heteropoly and Isopoly Oxometalates"; Springer-Verlag: New York, 1983; Inorganic Chemistry Concepts. (b) Weakley, T. J . R. Struct. Bonding 1974,18, 131. (c) Tsigdinos, G. Topics Curr. Chem. 1978, 76, 1. (d) Evans, H . T., Jr. Perspect. Struct. Chem. 1971, 4 , 1. (e) Kepert, D. L. Prog. Inorg. Chem. 1962, 4 , 199. (f) Tytko, K.-H., Glemser, 0. Adv. Inorg. Chem. Radiochem. 1976, 19, 239. (4) A large number of examples of polyoxoanions in catalysis, largely heterogeneous catalysis, exist. A few references in acid catalysis$" oxidation

catalysis,"" homogeneous Wacker-type4'J chemistry, and recent reviews4' are provided below. (a) Onoue, Y.; Mizutani, Y.; Akiyama, S.; Izumi, Y. CHEMTECH 1978,432 (b) Ono, Y.; Baba, T.; Sakai, J.; Keii, T. J . Chem. Soc., Chem. Commun. 1981,400. Baba, T.; Sakai, J.; Ono, Y. Bull. Chem. Soc. Jpn. 1982,55, 2657. (c) Okuhara, T.; Kasai, A,; Hayakawa, N.; Yoneda, Y.; Misono, M. J . Catal. 1983, 83, 121. (d) Hayashi, H.; Moffat, J. B. J . Catal. 1983.83, 192; Ibid. 1983, 81, 61. (e) Konishi, Y.; Sakata, K.; Misono, M.; Yoneda, Y. J . Catal. 1982, 77, 169. (f) Ai, M. J . Catal. 1981, 71, 88. (9) Akimoto, M.; Tsuchida, Y.; Echigoya, E. Chem. Lett. 1980, 1205. (h) Akimoto, M.; Tsuchida, Y.; Sato, K.; Echigoya, E. J . Catal. 1981, 7 2 , 83. (i) Ogawa, H.; Fujinami, H.; Taya, K. J . Chem. Soc., Chem. Commun. 1981, 1274. 6 ) Taraban'ko, V. E.; Kozhevenikov, I. V.; Matreev, K. I. Kinet. Katal. 1978, 19, 1160. (k) Kozhevnikov, I. V.; Matveev, K. I. Appl. Catal. 1983, 5, 135; Russ. Chem. Rev. (Engl. Transl.) 1982, 51, 1075.

0002-7863/84/ 1506-5750$01.50/0

2661 2671 26'81 m/z Figure 2. Calculated (solid line) vs. observed (dotted line) isotopic distribution patterns for the [M - 2H K]- (= KSHSiW9V,O4,-) ion a t m / z 2671. Only the centermost lines of the two patterns a r e directly comparable due to the presence of overlapping patterns in the observed F A B M S for the KSHSiW9V3040-ion.

+

this development, however, are well-known difficulties in obtaining accurate analytical8 and molecular weight data,3aproblems that (5) For photochemical applications, see: (a) Papaconstantinou, E. J . Chem. Soc., Chem. Commun. 1982, 12. (b) Yamase, T. Inorg. Chim. Acta 1983, 76, L25; Ibid. 1981,54, L207. (c) Yamase, T.; Sasaki, R.; Ikawa, T. J . Chem. Soc., Dalton Trans. 1981, 628. (d) Hill, C. L., unpublished results. Cited at the Proceedings of the Joint NSF-CNRS Polyoxoanion Workshop, St. Lambert des Bois, France, July 11-13, 1983.

0 1984 American Chemical Society